Open-source Zero-shot Explicit Binary BERT Text Classification Model - Achieve Binary Text Classification without Sample Training

Zero Shot Explicit Binary Bert

Developed by claritylab

A zero-shot text classification model based on BERT architecture, trained for binary classification on the UTCD dataset using an explicit training strategy

Text Classification

Transformers

EnglishOpen Source License:MIT #Zero-shot classification #Binary text classification #Explicit training strategy

Downloads 23

Release Time : 5/15/2023

Model Overview

A model specifically designed for zero-shot text classification tasks, trained for binary classification on the aspect-normalized UTCD dataset using an explicit training strategy

Model Features

Zero-shot learning capability

Can classify new categories without task-specific training data

Explicit training strategy

Uses a special training method to improve zero-shot classification performance

Binary classification framework

Focuses on binary classification tasks to provide more precise classification results

Model Capabilities

Zero-shot text classification

Binary text classification

Cross-domain text understanding

Use Cases

Natural Language Processing

Intent recognition

Identify intent categories in user text

High-accuracy intent classification

Content classification

Automatically classify text content

Classification capability without specific training data

🚀 Zero-shot Explicit Binary BERT

A BERT model designed for zero-shot text classification, offering efficient and accurate text categorization without prior training on specific labels.

🚀 Quick Start

This is a BERT model introduced in the Findings of ACL'23 Paper Label Agnostic Pre-training for Zero-shot Text Classification by Christopher Clarke, Yuzhao Heng, Yiping Kang, Krisztian Flautner, Lingjia Tang and Jason Mars. The code for training and evaluating this model can be found here.

✨ Features

Zero-shot Text Classification: Capable of classifying text into predefined categories without specific training on those labels.
Binary Classification Framework: Trained under a binary classification framework for enhanced performance.
Aspect-normalized Dataset: Utilizes the aspect-normalized UTCD dataset for training.

📦 Installation

Install our python package:

pip install zeroshot-classifier

💻 Usage Examples

Basic Usage

>>> from zeroshot_classifier.models import BinaryBertCrossEncoder
>>> model = BinaryBertCrossEncoder(model_name='claritylab/zero-shot-explicit-binary-bert')

>>> text = "I'd like to have this track onto my Classical Relaxations playlist."
>>> labels = [
>>>     'Add To Playlist', 'Book Restaurant', 'Get Weather', 'Play Music', 'Rate Book', 'Search Creative Work',
>>>     'Search Screening Event'
>>> ]

>>> query = [[text, lb] for lb in labels]
>>> logits = model.predict(query, apply_softmax=True)
>>> print(logits)

[[1.0987393e-03 9.9890125e-01]
 [9.9988937e-01 1.1059999e-04]
 [9.9986207e-01 1.3791372e-04]
 [1.6576477e-03 9.9834239e-01]
 [9.9990320e-01 9.6742726e-05]
 [9.9894422e-01 1.0557596e-03]
 [9.9959773e-01 4.0229000e-04]]

📚 Documentation

Model description

This model is intended for zero-shot text classification. It was trained under the binary classification framework via explicit training with the aspect-normalized UTCD dataset.

Finetuned from model: bert-base-uncased

📄 License

This project is licensed under the MIT license.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご